Scholarly Peer-Review via Token-Curated Registries*

In scientific publishing, the peer-review process is used to guarantee the quality of a particular publication, be it a journal article, a program for a scientific conference or a repository of data and code.

Many variants of the peer-review process exist. They all have the same goal: to assist the evaluation of whether a particular piece of research should be accepted into the publication in question.

Some Problems with Peer-Review

While peer-review has been employed in the past to great benefit of the scholarly community, several problems with peer-review exist:

1. Peer-review can be expensive. Well functioning peer-review processes are not easy to implement. They are time- and resource-consuming. As a result, effective peer-review is mostly employed in high-priced journals by international publishing houses.
2. Peer-review slows down research. Turn-around time for reviewing a piece of scholarship until its publication are often too long.
3. Peer-review moves scientific discourse behind closed bars. Comments by reviewers are often sent only to the editor and the author. The general readership will not see them.
4. Peer-review does not credit contributions well. Reviewers do not get sufficient credit for their work because they often remain anonymous and therefore their input cannot be sufficiently credited.

Token-Curated Registries as Blockchain-Based Peer-Review Systems

In this article, I propose to model the scholarly review-process as a token-curated registry (TCR). This blockchain-based model essentially casts the review-process as a protocol for finding consensus on whether a given research paper should be accepted into the publication. The basic mechanism for this process is a blockchain-based vetting process initiated by the researcher submitting a proposal to a journal and carried out by the token-holders of that journal.

What are Token-Curated Registries?

In its most basic form, token-curated registries are decentrally curated lists in which economic incentives of all participants are balanced by a carefully crafted token design and protocol.² TCRs produce lists of items that are of interest to a group of “consumers”. The owners of the list are the so-called “token holders”, while those that would like to be included in the list are “candidates”.

The Incentives of User Types in Token-Curated Registries

Each of the three user types has different economic incentives, but the important point is that their incentives are all aligned towards a common goal: creating a high-quality registry.

Consumers want to have a high-quality registry that helps them to make decisions, learn something new or use their time efficiently, for instance by not wasting their attention on irrelevant products, services and so on.

Candidates want to receive the attention of consumers and therefore would like to be included in the registry. The higher the quality of the registry, the higher the value for those candidates who will be included.

Token holders have the intrinsic motivation to produce a high quality list by providing thorough curation services. The higher the quality of the list, the more attention it will get from consumers, the more candidates will seek inclusion in the registry. These factors drive the value of the registry’s token.

Application, Challenge and Voting are the Central Mechanism for Curating a TCR

The central mechanism of a TCR is constituted by the application process for including a new item on the list. The process contains the following steps:

1. A candidate who wishes to be included in the registry, submits an application. As part of the application the candidate puts up a stake in the form of the native token of the registry (in TCR’s 1.0. Specification this is called the MIN_DEPOSIT parameter)
2. Token holders are informed of the new application and can review it within a given time frame (APPLY_STAGE_LEN). During this duration, a challenge on the inclusion of the candidate can be raised or if no challenge is raised, the candidate is included into the registry.
3. If an application gets challenged, a vote between all token-holders on the inclusion of the candidate is held. This vote is both token-weighted and secret (“commit-reveal”), the idea being that token-holders with more tokens have a higher natural incentive to curate high-quality lists than those with fewer tokens, and because open votes of token-heavy curators could influence votes of others.
4. If the application of the candidate was successful, the deposit of the candidate remains with the candidate and turns the candidate into a token-holder. If the application was successfully challenged, the deposit is forfeited and distributed to those who have challenged the application.

How can the TCR Model by Applied to Peer-Review Systems?

The entities in the TCR model can be easily mapped to describe a simple peer-review process. Given an implementation in a peer-reviewed journal, the mapping would look like this:

1. Token Holders are the primary care-takers of the journal. They include the reviewers, editors and publishers of the journal. These groups are formed from the pool of initiators and past authors of the journal.
2. Candidates are those wishing to publish a new article in the journal. As authors they would like to receive credit and recognition by having their work published in the journal.
3. Consumers are those who are reading the journal or who are vetting the quality of a scholar by looking at her list of contributions to the field.

Some Crypto-Economic Implications of the TCR Approach on Scholarly Peer-Review

1. TCR speeds up peer-review because the time-consuming mandatory discussion and evaluation is only needed in cases where an application is challenged.
2. Comments by reviewers during the challenge stage are going to be made public after voting and published along the article. If the article gets rejected, the comments will be either discarded or sent to the author.
3. Candidates who are also token holders (i.e. those candidates who have already been published in the journal) are excluded from voting. They also do not have the right to challenge their own work.
4. Token holders with their identity and with their amount of token are publically listed on the journal website. Depending on the amount of token they own, they can be designated publishers, editors-in-chief, editors or merely contributing authors.
5. Votes are also made public after the challenge process is completed to dis-incentivise clan-voting.
6. The journal is owned by the token holders. They are the ones contributing high quality research and those who successfully challenge work that is not of the right quality.
7. Cost for infrastructure can be deduced from the stakes of candidates and forfeited tokens from challenges. This way the journal can run without third party subsidies.
8. The system could have an arbitrary inflation of tokens, say 10%, for those token holders who do not participate in the review process. This way, the journal is fully community run.
9. To ensure widespread and early adoption of the journal, the first tokens can be airdropped among the community of interested scholars.

Get Best Software Deals Directly In Your Inbox

Conclusion

Token-curated registries could become a simple system for building peer-reviewed journals that are community-run and will result in high quality submissions. High-priced journal subscriptions will probably not disappear overnight, but with more and more high-quality journals running on token-curated registries, research will slowly but surely be made available to the public. One unsolved problem is how to seed initial token holders. What is the best mix to include so that a fair and balanced development of the registry can ensue?

* Thank you to Sven Merten and Simon de la Rouviere (@simondlr) for commenting on an early draft of this article. And thanks to Sven Merten for making the title much better. My special thanks go to Trent McConaghy for bringing TCRs to my knowledge, Mark Gotham for joining me in setting up the Journal of Computational Musicology and Simon de la Rouviere, Mike Goldin, Sven Merten, Oliver Völkel, Masha and Trent McConaghy for pushing the boundaries of my thinking as it relates to tokenization.

² The concept of TCRs has first been described by Mike Goldin in September 2017 (https://medium.com/@ilovebagels/token-curated-registries-1-0-61a232f8dac7). Since then, the specification has been further developed (https://medium.com/@ilovebagels/token-curated-registries-1-1-2-0-tcrs-new-theory-and-dev-updates-34c9f079f33d),resulting in a generic TCR repository for use in Ethereum applications (https://github.com/skmgoldin/tcr).